“Algorithmic trading made easy for everyone”
In the world of trading, Algorithms are computer programs that allow traders to execute trading activities automatically; i.e without the need for human interference. This attribute makes algorithms much faster in taking decisions and executing trades than the normal human ever could. [Pls note “algo” is just a shortened form of algorithms and I will be using algo throughout this casestudy]
Since the advent of trading algorithms in the 1980s, they have been a welcome development, but over the years traders began to point out some underlying issues restricting wide adoption. Some of these roadblocks include:
- Inability to write code
- The high cost of existing algorithms
- Absence of financial resources needed to run algorithms around the clock
- Difficulty in accessing already created algorithms
- …and a lot more intricate roadblocks.
Algopear is a startup dedicated to eliminating these and all the other unlisted barriers of entry into algorithmic trading. Algopear’s product is a platform that connects investors and traders to a library of ready-to-use algorithms that they could use to automate their trading activities.
When I joined the team at Algopear for this project, the company wasn’t called Algopear then. It was called Pushstash and the change of name to Algopear was a direct result of this massive overhaul I and the team executed and shipped over the course of approximately 6 months.
Some of the issues that prompted the redesign were:
- Low user retention
- Poor aesthetics
- Inability to convince users to take action
- Inability to meet the company’s financial goals
- We also needed to efficiently incorporate the new features conceptualized to open new streams of revenue for the company
–
Below are some screens from the old version of the app:
–
Problem statement
Besides Algopear there were other companies who were headed in similar directions and if Algopear was to dominate the industry and also stand the test of time as planned, then a solid design foundation was necessary. The challenge was to come up with efficient scale-able solutions to the identified problems and then translate these solutions into exciting easy-to-use interfaces and experiences. I also needed to create a design system that was scale-able and would be capable of accommodating new features in the future.
–
My role
Solo product designer
–
The team
Me [Designer], Bennie [Head of marketing], Lakeisha [Lead dev who’s also the co-founder], Ronnie [Founder], and about eight [8] other developers.
–
Project begins
For this project, I employed my four-step design process.
- Understand
- Design
- Test
- Iterate
–
Tools used:
Figma, Miro, Photoshop, Adobe illustrator, Google docs,Overflow.io, Xmind, Raindrop.io, and quite a few others.
–
Understand
“At the outset of the project, we didn’t really have a clear direction for the Algopear experience.”
In the old app, we simply sent users stock alerts on their preferred stocks and allowed them to monitor the performance of their favorite stocks. But in this new version, we wanted to do way more.
–
Instead of trying to figure out from the start what new features to add, I ensured we first answered questions like,
- What we wanted users to achieve from interacting with our platform and
- What users themselves say they want from the platform
- and How best we could evoke the correct emotions of safety, confidence, and trust throughout the experience.
–
Hypothesis statements — I started by working to identify all the things the team held as truths, especially the ones that hadn’t been confirmed through research. This was so I could work towards proving or disproving these opinions during research. Some of the hypothesis we held were:
- Users would be less hesitant to enter the world of algorithmic trading if we provide safeguards that protect them.
- Users would be more willing to try things out if we tailored the app’s experience to their trading style and risk appetite.
- Even in the midst of automation, users would still like to feel in control of the process.
- Users would like to see necessary financial stats in their money’s performance, quickly and effortlessly.
- Users would like their own privately curated algorithm lists separate from the marketplace lists
- Users would appreciate recommendations on algorithms that have performed well and are likely to perform well
–
Demographics — In collaboration with the marketing lead, I identified the three major demographic categories our existing and potential user base would fall under. I named them D1, D2, and D3.
D1: Investment savvy people, high-risk appetite, very financially capable and experienced folks. High risk – High rewards kind of traders.
D2: Average experience with high financial capability. Averagely experienced with average financial capability. Mid risk – Mid rewards kind of trader. Average risk appetite too.
D3: Low-risk appetite, little to no experience. Very unwilling to lose money. Not as financially capable as the other categories.
–
Business goals — From the meetings and discussions we had, I was also able to deduce some business goals. This part of the process is essential because keeping these in mind help inform certain decisions to be taken later on in the design process
Short-term:-
- Secure another round of funding
- Build an app that would give a solid experience all-round
- Grow to first hundred thousand users
- Begin the path to becoming a household name in the trading space
- Communicate credibility and expertise in every interaction
Mid-term:-
- Induce profits from algorithm sales and subscriptions
Long-term:-
- Dominate the retail algorithm industry
- Educate more people on trading and the markets and basically demystify the financial markets to more people
KPIs — The major KPI was an increase in user base on both the algo maker and trader side.
–
User research recruitment — For interviews, I was looking for people who fit our three different demographics. I sent out initial questions via a questionnaire and then I reached out to users with interesting responses to gain further insight. Further discussions allowed me to uncover new insights and also invalidate some of the hypotheses we held.
User research findings — After hearing from and discussing with potential users, Some common themes began to emerge. While some of our hypotheses were confirmed, I also discovered many things we were missing.
- New or Low experienced users agreed that they actually would like to start in a way that posed the least risk and work their way up from there.
- Even though algorithms were automated and could pause transactions by themselves, users still wanted to be able to shut down things ASAP whenever they felt like it. — *Now even though this isn’t an action we wanted users to always take as it could be done at wrong times, we decided we definitely needed to give them the feature. But to also satisfy what we felt was the right thing to do, I came up with a solution which was to provide, infobox UI elements where we would advise users on whether or not to take certain actions, but ultimately the decision still rested in their hands.
- Users agreed they would like to be able to compare and contrast the stats of different algorithms to decide which was best for them but had differing opinions on what stats they paid attention to compared to what we thought they paid attention to
- Experienced users also pointed out that they wanted to be able to analyze things quickly hence needed any stats page to be properly laid out but they wanted emphasis on certain metrics.
–
My research also birthed some “How might we” statements like
- How might we eliminate the resistance people have to anything involving trading
- How might we make the experience engaging and pleasurable
- How might we demonstrate our unique value proposition best
- How might we ensure the entire experience is free of difficulty
- How might we make the app’s experience easy for inexperienced traders
- How might we bring users up-to-speed on difficult terms
–
Design
High-FI Designs
- User profile – dedicated to showing the picture of the active user and also acting as a button to jump into the profile settings of the app
- Timeframe switcher – This is an essential part of the home screen because all the information below it shows based on what timeframe the user is looking at
- User’s balance -This entire section shows how much the user has left and also shows the percentage increase or decrease within the selected timeframe above
- Graphical representation of money stats – This section shows with bar charts a visual representation of the amount of money spent by the algorithm and the amount of profit gained or loss incurred. I went with bar charts because this information is the kind where users need to quickly in one glance see what periods of the day/month/year the algorithm spent low amounts and high amounts
- Performance info – This is the infobox idea I talked about earlier, this design solution of mine is used throughout the app with the aid of color states to advise, warn or commend users.
- Recent algo activity – This section shows the two most recent algorithm activities with the option to jump into the specified algo page to see all of its activity in realtime
- Stats minimizer – As you can see this homepage is really long, from talking to users I realized that the users didn’t really dislike it but wanted it to be shorter if possible and without taking away anything from the homepage because they still wanted to see those stats, so my solution was another design solution that allowed users minimize sections of a page, so basically everything above that button is minimizable and then the home screen looks like the version beside it and of course the minimized version comes with the option to expand things back into the previous view.
- To-dos info – The to-dos screen is a compilation of tasks users need to take action on. It comes with clear explanatory text and the individual elements upon click initiate the process of completing the task.
- Recent Algo lists – From the marketplace users can save algorithms into lists of their own. I incorporated this feature for two reasons, First is based on my previous e-commerce experience, I know that when in a market place users like to collect things that interest them into a separate place that they can then review later and make their decision on which to purchase. Secondly, users explained during research that they would like to curate their own algorithm lists and share them with others, especially experts who had a following that most times consisted of less experienced traders who trusted their judgment.
- Algorithm type switcher – This element is the highest level and most used filter option. This I discovered from users and testing. So I decided to reduce the number of clicks needed for this quick sorting by bringing out this section which was previously tucked away inside the filter screen and bring it out and make it prominent and much easier to reach.
- Single algorithm – This element represents a single algorithm on the market, According to users, I learned that when deciding which algorithms to even consider, that the two most useful stats they wanted to see were how old the algorithm was and how successful it was, so I attached these two pieces of information to individual algorithm elements so they could be easily seen without having to click inside each algorithm first.
- Featured algorithm _ This is a variation of an individual algorithm on the marketplace with a star on it to indicate that it’s being given prominence aka is being featured based on a consistently good performance.
- Quick add to list button – This is another solution that was born from testing insights. So like I said users wanted to be able to add algorithms to their own lists so they can go over them later, but simply giving them lists wasn’t enough, I needed to also incorporate a swift way of adding marketplace items to this list and that was how this idea came about, it’s a simple hold and swipe that reveals a plus button that can be used to quickly add an item to a list.
- Important algorithm stats – Inside an individual algos detail page, I sow the four most important stats users informed us they wanted to see, actually I was six stats that got mentioned the most but from talking some more with both the users and the founders we decided on four ut of the six that were of utmost importance and we went with those.
- Timeframe switcher
- Performance stats graph
ITERATE
Testing results
Deliverables
Results & Impact